linux_dsm_epyc7002/arch/tile/kernel
Chris Metcalf 1337173148 arch/tile: fix __ndelay etc to work better
The current implementations of __ndelay and __udelay call a hypervisor
service to delay, but the hypervisor service isn't actually implemented
very well, and the consensus is that Linux should handle figuring this
out natively and not use a hypervisor service.

By converting nanoseconds to cycles, and then spinning until the
cycle counter reaches the desired cycle, we get several benefits:
first, we are sensitive to the actual clock speed; second, we use
less power by issuing a slow SPR read once every six cycles while
we delay; and third, we properly handle the case of an interrupt by
exiting at the target time rather than after some number of cycles.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
2011-03-01 16:20:04 -05:00
..
asm-offsets.c
backtrace.c arch/tile: support new info op generated by compiler 2010-10-15 15:39:25 -04:00
compat_signal.c arch/tile: handle rt_sigreturn() more cleanly 2010-12-17 16:59:29 -05:00
compat.c BKL: remove extraneous #include <smp_lock.h> 2010-11-17 08:59:32 -08:00
early_printk.c arch/tile: bomb raw_local_irq_ to arch_local_irq_ 2010-11-01 15:30:42 -04:00
entry.S arch/tile: fix __ndelay etc to work better 2011-03-01 16:20:04 -05:00
hardwall.c arch/tile: mark "hardwall" device as non-seekable 2010-11-01 15:31:42 -04:00
head_32.S arch/tile: catch up with section naming convention in 2.6.35 2011-03-01 16:18:52 -05:00
hvglue.lds
init_task.c
intvec_32.S arch/tile: handle rt_sigreturn() more cleanly 2010-12-17 16:59:29 -05:00
irq.c tile: Use proper accessor functions in show_interrupt() 2011-02-23 16:07:34 -05:00
machine_kexec.c arch/tile: complete migration to new kmap_atomic scheme 2010-11-01 15:30:36 -04:00
Makefile pci root complex: support for tile architecture 2010-11-24 13:13:49 -05:00
messaging.c arch/tile: bomb raw_local_irq_ to arch_local_irq_ 2010-11-01 15:30:42 -04:00
module.c
pci-dma.c
pci.c pci root complex: support for tile architecture 2010-11-24 13:13:49 -05:00
proc.c
process.c arch/tile: bug fix: exec'ed task thought it was still single-stepping 2011-03-01 16:19:58 -05:00
ptrace.c arch/tile: don't allow user code to set the PL via ptrace or signal return 2010-11-01 15:31:17 -04:00
reboot.c arch/tile: bomb raw_local_irq_ to arch_local_irq_ 2010-11-01 15:30:42 -04:00
regs_32.S arch/tile: parameterize system PLs to support KVM port 2010-10-15 15:38:09 -04:00
relocate_kernel.S
setup.c arch/tile: make glibc's sysconf(_SC_NPROCESSORS_CONF) work correctly 2010-11-24 13:42:15 -05:00
signal.c arch/tile: handle rt_sigreturn() more cleanly 2010-12-17 16:59:29 -05:00
single_step.c arch/tile: bug fix: exec'ed task thought it was still single-stepping 2011-03-01 16:19:58 -05:00
smp.c arch/tile: bomb raw_local_irq_ to arch_local_irq_ 2010-11-01 15:30:42 -04:00
smpboot.c BKL: remove extraneous #include <smp_lock.h> 2010-11-17 08:59:32 -08:00
stack.c
sys.c BKL: remove extraneous #include <smp_lock.h> 2010-11-17 08:59:32 -08:00
tile-desc_32.c
time.c arch/tile: fix __ndelay etc to work better 2011-03-01 16:20:04 -05:00
tlb.c
traps.c arch/tile: enable single-step support for TILE-Gx 2010-10-15 15:38:26 -04:00
vmlinux.lds.S arch/tile: catch up with section naming convention in 2.6.35 2011-03-01 16:18:52 -05:00